Artificial Intelligence Engineer

Roto Pumps; Dadri; On-site; INR 1,200,000 - 2,400,000; Full time

Job summary

Roto Pumps is building an enterprise-grade AI platform to support engineering, sales, and marketing teams with self-hosted and API LLMs, RAG, and autonomous agents. You will ship production AI systems across LLMs, RAG pipelines, and document intelligence.

You will work on vector databases, document ingestion, and GPU-based model serving while integrating AI with ERP/CRM workflows in a manufacturing environment.

Responsibilities

• Build and maintain enterprise AI applications and assistants for engineering, sales, marketing, procurement, and support.
• Build RAG systems on company knowledge bases.
• Create and maintain vector databases (Qdrant), with embedding and reranking models.
• Tune retrieval pipelines for accuracy and evaluate answer quality.
• Build autonomous, multi-step AI agents using LangGraph or similar.
• Integrate agents with internal systems (ERP, CRM) via MCP servers.
• Integrate self-hosted and API LLMs (vLLM, OpenAI, Claude, Gemini).
• Optimize prompts; apply safety controls (NeMo Guardrails).
• Deploy with Docker; monitor with observability tooling.

Skills

LLM Integration

RAG Systems

Vector Databases

API Development

Docker & GPU Serving

Observability Tools

Tools

Docling

LlamaParse

Qdrant

vLLM

LangGraph

MCP servers

Job description

Roles & Responsibilities:

AI Platform & RAG

• Build and maintain enterprise AI applications and assistants for engineering, sales, marketing, procurement, and support.
• Build RAG systems on company knowledge bases.
• Build document ingestion pipelines for PDFs, manuals, brochures, and specs using tools like Docling and LlamaParse.
• Create and maintain vector databases (Qdrant), with embedding and reranking models (e.g. BGE-M3, BGE-Reranker).
• Tune retrieval pipelines for accuracy and evaluate answer quality (e.g. RAGAS).

Agentic AI

• Build autonomous, multi-step AI agents using LangGraph or similar.
• Integrate agents with internal systems (ERP, CRM) via MCP servers.
• LLM Integration
• Integrate self-hosted (DeepSeek, Qwen via vLLM) and API LLMs (OpenAI, Claude, Gemini, OpenRouter).
• Optimize prompts; apply safety controls (e.g. NeMo Guardrails).

Preferred candidate profile

We are building a self-hosted, enterprise-grade AI platform for our manufacturing, engineering, sales, and marketing teams. We need a hands-on AI Engineer to help build and maintain it working across LLMs, RAG, autonomous agents, vector databases, document intelligence, and GPU-based model serving, using AI coding assistants like Claude Code and Codex to move fast. This is a builder's role: you will ship production systems, not prototypes.

Workflow Automation

• Build business automation in n8n.
• Connect AI with ERP, CRM, email, WhatsApp, and internal apps.

Backend & Infrastructure

• Build REST APIs with FastAPI.
• Deploy with Docker and Docker Compose; work with GPU-based model serving (vLLM).
• Monitor with observability tooling (Langfuse, Grafana, Prometheus).